[MITgcm-support] Convert data in latlon grid to cube-sphere grid
dmenemenlis at gmail.com
Tue Dec 23 09:26:29 EST 2008
Yuan, in MITgcm/utils/matlab/cs_grid there is a bunch of matlab
functions that are specific to cube-sphere grid. In particular,
uvLatLon2cube.m will convert vectors from a lat/lon grid to cube sphere.
For surface forcing fields, I suggest that you use exf_interp for on-
the-fly interpolation. An example is in MITgcm_contrib/high_res_cube/
I use bi-cubic interpolation for vector fields to avoid artifacts in
the gradients and bi-linear for everything else to avoid spurious
On Dec 23, 2008, at 5:09 AM, Yuan Lian wrote:
> Hi Dimitris,
> I talked a little about this with Jean-Michel during AGU. I went to
> check your poster but you had left.
> Anyway, I do need to convert the vector field as well. For example,
> I want to use wind field from
> observation data as initial condition. Will the bilinear
> interpolation cause any trouble here (e.g. cell centered
> to a standard C-grid)? I am thinking about making some matlab
> scripts to do the trick. Any hint here?
> Thanks a lot for your help!
> Dimitris Menemenlis wrote:
>> Yuan, did anybody get back to you? Going from latlon to cube
>> sphere for a scalar field is easy. You can just use bilinear
>> interpolation, for example, function "interp2" in matlab. For
>> vector fields it's a bit trickier but MITgcm can do that on the
>> fly. Dimitris
>> On Dec 3, 2008, at 7:17 PM, Yuan Lian wrote:
>>> I am trying to generate topography files in cube-sphere grid. The
>>> original data
>>> is in latlon grid. More generally, is there anyway I can convert
>>> the data in latlon grid to
>>> cube-sphere grid? Are there any Matlab scripts available? Thanks!
>>> MITgcm-support mailing list
>>> MITgcm-support at mitgcm.org
>> MITgcm-support mailing list
>> MITgcm-support at mitgcm.org
> MITgcm-support mailing list
> MITgcm-support at mitgcm.org
More information about the MITgcm-support